$ loading_
Scan AI agents for tool-calling vulnerabilities and surface key security risks.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "Agent Security Scanner" MCP server from askskill: Run: claude mcp add 'io-github-mikehzp-agentvuln' -- npx -y agentvuln
Scan this AI agent's tool-calling chain for prompt leaks, tool hijacking, prompt injection, and unauthorized tool use. Return findings by severity with remediation advice.
A severity-ranked vulnerability report with risk explanations, affected steps, and remediation guidance.
Run a pre-release security scan on this agent. Simulate malicious inputs and abnormal tool responses to check for system prompt leaks, hijacked instructions, or injection acceptance.
A pre-release security assessment showing whether high-risk vulnerabilities exist and what should be fixed first.
Re-scan the patched agent and compare vulnerabilities before and after the fix to confirm prompt leaks, injections, and tool misuse issues are resolved.
A fix verification report showing resolved, unresolved, and newly introduced security issues.
No documentation provided
Check the source repo for usage and examples.
Scan AI agents, MCP servers, and skills for security risks.
Scan AI agent code for security issues from MCP-compatible clients.
Scan agent-generated code for vulnerabilities, prompt injection, and semantic security issues.
Scan AI agent skill files for credential theft and prompt injection threats.
Scan Claude Code configs for vulnerabilities, misconfigurations, and prompt injection risks.
Scan AI skills before install for injections, leaks, and supply-chain risks.